Data centre race: Singapore betting on quality over scale, say observers

Data centre race: Singapore betting on quality over scale, say observers


As the AI race heats up, the city-state is pursuing a ‘sustainable’ data centre strategy – but what stands in the way of its ambitions?

[SINGAPORE] As South-east Asia and the world bet on artificial intelligence (AI), data centres have become one of the hottest investments.

A November report from Fortune Business Insights estimated the global data centre market size at about US$243 billion in 2024, with projected growth to reach about US$270 billion in 2025. Much of this growth is expected to be driven by AI.

Meanwhile, data centre investments could hit US$7 trillion by 2030, with more than half dedicated to computing hardware, according to McKinsey and Company. About 60 per cent of that will be spent outside the US.

Singapore is keen to embrace AI and capture its economic benefits. In 2023, it unveiled the latest version of its National AI Strategy, which includes the 70-odd data centres in the country that are required for AI-related use.

The country has also pushed for “sustainable” data centres since lifting a three-year moratorium on new data centre projects in 2022. 

This year, the city-state announced plans to build a low-carbon data centre park on Jurong Island with up to 700 megawatts (MW) of capacity – its largest such facility to date.

The Business Times examines where Singapore’s data centre ambitions are headed in 2026, where the country stands in the regional data centre race, and the challenges it faces.

What Singapore wants

The Republic’s data centre strategy is carefully designed, say industry players and analysts.

Xavier Lee, equity analyst at Morningstar, notes that Singapore has demonstrated a “measured” yet “ambitious” push for quality over quantity. He cites the 20-hectare Jurong Island data centre park, which will increase Singapore’s data centre capacity by 50 per cent from 2024 levels, as an example.

Meanwhile, Keppel’s chief digital officer Manjot Singh Mann believes that Singapore has the potential to become a “very significant” digital hub to run the most critical and expensive AI applications. The country’s strengths in policy, execution and connectivity can be applied beyond AI training to more intelligent workloads, such as those deployed in banking or manufacturing, he notes.

“Training loads require much higher (data centre) capacity and much larger pieces of land,” says Mann.

On the other hand, intelligent workloads need to be closer, monitored and more secure. “Those workloads require a hub, and that hub is Singapore,” he adds.

Lee echoes this view, pointing to Singapore’s unique position of not having to directly compete with other Asean nations for data centres.

“Investors should view the region as a single, symbiotic ecosystem. Singapore serves as the brain for latency-sensitive, high-value workloads, while Johor and Batam provide the industrial scale for heavy AI model training,” he says.



Read Full Article At Source